Single precision better than double precision?

140raiders · July 30, 2015, 02:08

I am running a basic simulation with water flowing at about 0.3 m/s through a 20 m pipe. I have very small protrusions embedded in the pipe that are only a few millimeters in height, width, and length. I am only interested in the pressure drop along the pipe and I get very different results depending on if I use single precision or double precision. I am new to Ansys but I am assuming that the double precision results are more accurate and I should just use those. Am I correct in thinking this or could the single precision results actually be more accurate? Also, should double precision always be used so long as the longer computational time is of no concern?

Thanks in advance for the help!

ghorrocks · July 30, 2015, 03:32

Double precision will always be more accurate. If single precision is more accurate then this is simply a one-off chance as random noise has tried to confuse you.

But double precision requires more memory and a longer simulation time (assuming convergence is the same as a single precision simulation). Whether the longer simulation time and additional memory requirements are a problem is up to up. At least you have the choice.

July 30, 2015, 02:08	Single precision better than double precision?	#1
140raiders New Member Mike Join Date: Jul 2015 Posts: 1 Rep Power: 0	I am running a basic simulation with water flowing at about 0.3 m/s through a 20 m pipe. I have very small protrusions embedded in the pipe that are only a few millimeters in height, width, and length. I am only interested in the pressure drop along the pipe and I get very different results depending on if I use single precision or double precision. I am new to Ansys but I am assuming that the double precision results are more accurate and I should just use those. Am I correct in thinking this or could the single precision results actually be more accurate? Also, should double precision always be used so long as the longer computational time is of no concern? Thanks in advance for the help!

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
CFD by anderson, chp 10.... supersonic flow over flat plate	varunjain89	Main CFD Forum	18	May 11, 2018 08:31
Continuing User Defined Real Gas Model issues	aeroman	FLUENT	6	April 8, 2016 04:34
Missing math.h header	Travis	FLUENT	4	January 15, 2009 12:48
what's wrong about my code for 2d burgers equation	morxio	Main CFD Forum	3	April 27, 2007 11:38
REAL GAS UDF	brian	FLUENT	6	September 11, 2006 09:23

July 30, 2015, 03:32		#2
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,854 Rep Power: 144	Double precision will always be more accurate. If single precision is more accurate then this is simply a one-off chance as random noise has tried to confuse you. But double precision requires more memory and a longer simulation time (assuming convergence is the same as a single precision simulation). Whether the longer simulation time and additional memory requirements are a problem is up to up. At least you have the choice.